Haven’t tested that yet, but the Face_AnimBP has a setup where is getting the body animation from the main skeleton, and then it applied the facial animation blending the wo together.
It uses a Layered blend per bone node, which is using 3 joints, as you see in the image below
I think that, in order for the neck animation to work from the data from the iPhone, you need to use the neck_01 joint, instead of the ones you see in the Layered node.
By doing so, you get the animation from the iPhone from the neck_01 up, while the rest of the body is driven byt the body animation.